By way of brief background, conventional analysis of radio access network (RAN) conditions is often based simply on manipulations of current RAN performance values. These conventional techniques generally ignore historically similar performance. Moreover, conventional techniques can be considered rudimentary, for example, flagging a RAN based on a current performance indicator exceeding a threshold value. As such, conventional analysis of RAN performance can be improved upon, which can result in, for example, a reduction in false alerts to a underperforming RAN that can, in turn, reduce costs associated with responding to a false alert.